When you consider a tapeworm or an Ebola virus, it is easy think of them as being evil to their very core. That's a mistake. It's true that at this point in their evolutionary history these species have become well adapted to living inside of other organisms (us), and using our resources to help them reproduce themselves even if we get sick in the process. But one of the big lessons of modern biology is that there are no essences in nature--only the ongoing interplay of natural selection and the conditions in which it works. If the conditions change, organisms may evolve into drastically different things. Even the most ruthless parasite may discover the virtues of peace and harmony--if the conditions are right.

Joel Sachs and James Bull, two biologists at the University of Texas, have offered a vivid demonstration of this fact with the help of bacteria-infecting viruses, called bacteriophages. Bacteriophages, such as the one shown here, are wickedly elegant in the way they find hosts and inject their DNA, which then hijacks the bacteria's cellular machinery to make new bacteriophages. (For more of my praise of the bacteriophage, plus an excellent movie of the beast, go here.)

Bacteriophages fit the definition of parasite to a T. In many cases new viruses multiply inside a host until the bacterium simply rips apart. In other cases, they make bacteria sick, draining resources from their hosts that could otherwise be used for the hosts' own reproduction. But, as Bull and his colleagues have shown in a series of experiments, bacteriophages are not malicious in their very essence. Depending on the conditions in which bacteriophages find themselves, they can evolve into milder forms, or into meaner ones.

Bull and his colleagues took advantage of the fact that many bacteriophages can infect new hosts in one of two ways--by escaping one bacterium to invade another, or by getting passed down from one bacterium to its offspring. These two routes are called horizontal and vertical transmission. Bull's team experimentally created conditions that favored vertical transmission, and within a few dozen generations the viruses became much milder. If you rely on your host's survival for your own survival, it doesn't pay to be a brutal killer. (I wrote more about this evolutionary trade-off--and some of the debate surrounding it--in this article for Science.)

Now Bull and Sachs show that bacteriophages can even evolve to be nice to other bacteriophages. They describe the experiment in the January 11 issue of the Proceedings of the National Academy of Sciences. They started out with two bacteriophages, called f1 and IKe. Both viruses infect E. coli bacteria, but they enter in different ways. f1 only grabs onto one type of hair on the surface of E. coli (the F pilus), while IKe invades its hosts through another type (the N pilus). In the wild, f1 and IKe don't get along well. If they end up in the same host, they compete for the bacterium's cellular machinery. Also, because they are close relatives, sharing the same 10 genes, DNA-binding proteins of one bacteriophage can accidentally grab the DNA of the other species. As a result, bacteria infected with both f1 and IKe produce fewer copies of each virus than bacteria infected with only one species. It's the classic Darwinian scramble.

But Bull and Sachs wondered what would happen if the survival of both bacteriophages actually depended on their coexistence. Here's how they answered the question. First they engineered both bacteriophages, adding a gene that provides resistance to a different antibiotic (kanamycin for IKe and chloramphenicol for f1). Then they dumped billions of the engineered viruses into beakers full of E. coli. They allowed the viruses 16 minutes to find hosts, invade them, and start producing the proteins that confer antibiotic resistance. Then they added the two antibiotics to the beakers. Only the bacteria that had been infected with both bacteriophages could survive the assault. If a bacterium harbored only f1, for example, it would still die, because it remained susceptible to kanamycin.

Next, Bull and Sachs let the bacteriophages and their hosts alone for an hour. The bacteria divided, while the bacteriophages made copies of themselves. After an hour, the scientists dissolved away the bacteria, leaving behind the viruses. These new viruses were then added to a fresh batch of bacteria, and the cycle repeated itself.

Viruses are notoriously sloppy at replicating. The odds of a new virus winding up with a mutation is much higher than for organisms like ourselves, equipped as we are with enzymes that act like genetic proofreaders. As a result, with each round of Bull and Sachs' experiment, many variants emerged in both the f1 and IKe populations. The variants that were best suited for reproducing in the experimental conditions were favored by natural selection, and over time the viruses evolved. After 50 rounds, Bull and Sachs stopped the experiment and took a look at what the bacteriophages had become. Were they so selfish that they had driven themselves extinct? Or had they come to some sort of accommodation?

The bacteriophages clearly went through natural selection during just 50 rounds. By that point f1 was producing 50 times more copies of itself, and IKe was producing 1,000 times more. At the beginning of the experiment sharing a host was a bad thing for these viruses, but at the end it had become a very good thing. Bull and Sachs discovered that they had overcome their conflict of interest in an extraordinary way: they practically merged into a single organism. When Bull and Sachs opened up a bacteriophage shell, very often they found both the f1 and IKe genomes sitting side by side. They cold still find plenty of viruses with a single genome inside, but even in these cases, evolution had taken a dramatic turn. By about round 20, the IKe viruses had lost the ability to make their own protein coat. Instead, they borrowed f1 coats.

Bull and Sachs argue that the bacteriophages adapted to the experiment in a clever way. If you're a bacteriophage, successfully invading a host on your own is not enough to stave off death, because you may find yourself alone. If a mutation lets you bring along the other virus with you, then you are pretty much guaranteed survival. For some reason, f1 seems to have taken the lead in this cooperation, mutating in such a way that IKe genomes could slip easily inside f1's protein coats. As a result, IKe began to lose its own ability to survive as an independent virus, relying instead on the cooperation of f1. Once the viruses were packaged together, they no longer had a conflict of interest, and they could evolve an even greater level of cooperation.

Evolutionary biologists have long been fascinated by cooperation, whether the cooperators are chromosomes in a single cell, individual bacteria in a colony, or people in a village. What keeps individuals from cheating on others, from choosing the selfish strategy rather than the selfless one? Scientists have constructed sophisticated mathematical models in order to find the right sort of conditions where cooperation might evolve. But Bull and Sachs point out that it only took them 50 generations to turn uncooperative bacteriophages into intimate partners. When they sequenced the viruses, they found that f1 had acquired just eight mutations in its DNA, and IKe had acquired nine. Perhaps cooperation is not such a big deal after all. And perhaps parasites are not the essence of evil we tend to believe them to be.